Architecture Modification / Research · Jupyter Notebook
VISTA
VISTA: Velocity-Informed Staticness Temporal Attention A Physics-Aware Motion Prior for Stable Object Detection in Automotive Perception
What it is
VISTA is a video object detection framework that augments standard detection architectures (DETR, YOLO, SSD) with optical flow-based motion features and a staticness branch to improve detection of moving and static objects. It provides training scripts for three variants and includes inference code.
Where it counts
- Machine Learning90%
- Computer Vision90%
- Data Science20%
How much of the project is about each field, as scoring counts it. It ranks in the field it is most about.
Around it on the Machine Learning board
Whole board- 33AgroADASML Application · Python52.6ImplementationSuceveanu Mircea Dan SanduMachine Learning
- 34ml-from-scratchEducational Implementation · Jupyter Notebook51.9ImplementationPop AlexandruMachine Learning
- 35VISTAArchitecture Modification / Research · Jupyter Notebook50.4ImplementationSuceveanu Mircea Dan SanduMachine Learning
- 36chatbot-pipelineML Application · Python50.2ImplementationPop AlexandruMachine Learning
- 37socrateLibrary / Framework · Jupyter Notebook49.2Implementationvlad andreiMachine Learning